Prostate cancer is the most common cause of cancer in men. In 1996, 317,000 new cases of prostate adenocarcinoma were diagnosed and over 41,400 men died of the disease (Karp et al., 1996). Only lung cancer has a higher mortality. The chance of a man developing invasive prostate cancer during his lifetime is 1 in 6 or 15.4%. At the age of 50, a man has a 42% chance of developing prostate cancer and 2.9% of dying from the disease. While advances in early diagnosis and treatment of locally confined tumors have been achieved, prostate cancer is incurable once it has metastasized. Patients with metastatic prostate cancer on hormonal therapy will eventually develop an androgen-refractory (androgen independent) state that will lead to disease progression and death.
The major cause of morbidity and mortality from prostate cancer is the result of androgen-independent metastatic tumor growth. As a result, there is great interest in defining the molecular basis for advanced staged disease with the hope that these insights may improve the therapeutic options for these patients. However, progress in this area has been difficult for a number of reasons. For example, the availability of prostate tissue for molecular studies is limited because most prostate tumors are small. Moreover, there is tremendous heterogeneity within surgical prostatectomy tumor samples, it is difficult to reducibly culture prostate cancer explants in vitro, and there are a limited number of immortalized prostate cancer cell lines.
There is, therefore, an interest in finding alternative procedures which will allow for stable growth of prostate cancer tissue, which in turn would allow for the investigation of the progression of prostate cancer in vivo, provide a stable supply of prostate cancer tissue and provide a model for metastatic expansion of prostate cancer which accurately simulates or mimics the biology of the disease.
There is also a need for more reliable and informative staging and prognostic methods in the management of advanced prostate cancer. Clinically staging prostate tumors relies on rectal examination to determine whether the tumor remains within the borders of the prostatic capsule (locally confined) or extends beyond it (locally advanced), in combination with serum PSA determinations and transrectal ultrasound guided biopsies. However, none of these techniques has proven reliable for predicting progression of the disease.
The primary sites of prostate cancer metastasis are the regional lymph nodes and bone. Bone metastases occur in sites of hematopoietically active red bone marrow, including lumbar vertebral column, ribs, pelvis, proximal long bones, sternum and skull. Bony metastases of prostate cancer differ from those of other tumors that commonly colonize in bone in that they are characterized by a net gain in bone formation (osteoblastic) rather than resorption predominant in bone metastases of breast cancer and melanoma.
Until recently, bone metastasis was thought to be a late stage in disease progression. However, the recent development of highly sensitive techniques (such as RT-PCR for prostate specific genes) to detect prostate cancer cells has revised this notion. Prostate cancer cells have been detected in the peripheral blood and bone marrow of patients with advanced stage disease using RT-PCR assays for PSA mRNA (Ghossein et al., 1995; Seiden et al., 1994; Wood et al., 1994; Katz et al., 1994) or immunomagnetic bead selection for PSA protein (Brandt et al., 1996). When positive, these tests show that prostate cancer cells represent about 0.1-1.0% of the circulating blood cells. Moreover, it is now clear that small numbers of prostate cancer cells circulate in the peripheral blood and lodge in the bone marrow even in patients with early stage, low risk disease (Olsson et al., 1997; Deguchi et al., 1997; Katz et al., 1996). Interestingly, these cells tend to disappear in most patients following radical prostatectomy (Melchior et al., 1997). These results suggest that the primary tumor site is a constant source for seeding the marrow, and that only a small subset of these cells have the capacity to grow into a metastatic lesion. This concept is consistent with estimates from animal models for other tumor types that only about 1 in 10,000 circulating cancer cells are able to lodge in and productively colonize other organs (Fidler et al., 1990).
The factors involved in advanced prostate cancer progression to bone metastasis are poorly defined. Anatomic, local bone/marrow and tumor cell factors are all believed to play a role. Baston described the extensive vertebral venous system that consists of a network of longitudinal, valveless veins that run parallel to the vertebral column and form extensive, direct anastomoses with the veins of the ribs, pelvis and brain (Baston, 1942). Prostate cancer cells entering prostatic veins may be transported via this plexus directly to these organs without entering the inferior vena cava of passing through the lungs. This hypothesized mechanism of metastasis both by clinical documentation of patterns of prostate cancer metastasis compared to other tumors and by animal models wherein occlusion of inferior vena cava during tail vein injection of tumor cells increased the incidence of vertebral metastasis (Nishijima et al., 1992; Coman and DeLong,1951).
Although the vascular anatomy is an essential component of the spread of prostate cancer to bone, it cannot fully explain the selective pattern of all skeletal metastases. Bone, which receives 5-10% of the cardiac output, is a more frequent metastatic site than would be expected from blood-flow criteria (Berettoni and Carter, 1986). Bone marrow consists of two clearly identifiable components: the hematopoetic cells which comprise the majority of the cellular elements, and stromal component that is formed of highly vascular connective tissue. The hematopoetic cells are transient in the bone marrow; upon maturation they move into the blood stream. The stroma, however, remains and serves as a scaffolding upon which the hematopoetic cells can differentiate and mature. One of the important factors in prostate cancer cells arresting in these sites is likely their adhesion to the bone marrow stroma. It has been demonstrated both in vitro and in vivo that tumor cells will preferentially adhere to the stromal cells of the organs to which they metastasize (Haq et al., 1992; Netland and Zetter, 1985; Zetter et al., 1992). When rat prostate cancer (MatLyLu) cells were injected into the left ventricle of syngenic rats, vertebral body metastases developed; these metastases were then collected, disaggregated and reinjected. Cell lines established after 6 similar passages through animals adhered strongly and preferentially to bone marrow stroma and endothelial cells (Haq et al, 1992). A similar approach has increased the incidence of metastasis from the LNCaP prostate cancer cell line in immune deficient mice (Thalmann et al., 1994).
It is critical that appropriate in vivo models for prostate cancer bone metastasis be developed to more fully explore the mechanistic aspects of this process. To date, most work in this area has focused on three human prostate cancer cell lines--PC-3, DU-145, and LNCaP (Lee et al., 1993). All three grow a subcutaneous nodules in immune deficient mice, and sublines with variable metastatic properties have been derived (Shervin et al., 1988, 1989; Wang and Sterarns, 1991; Kozlowski et al., 1988). However, none of these sublines has been shown to reproducibly give rise to osteoblastic lesions typical of prostate cancer. A major limitation of the DU-145 and PC-3 cell lines is the lack of prostate specific antigen (PSA) and androgen receptor (AR) expression (Kaighn et al., 1979; Gleave et al., 1992), which raises regarding relevance to clinical prostate cancer. The LNCaP cell line is androgen responsive and expresses PSA, but contains a mutation in the androgen receptor which alters ligand specificity.